Robust Intelligence's mission is to eliminate AI Risk. As the world increasingly adopts AI into automated decision processes, we inherit great risk.
Our flagship product is built to be integrated with existing AI systems to enumerate and eliminate risks caused by unintentional and intentional (adversarial) failure modes. With Generative AI becoming increasingly popular, new vulnerabilities and attacks present a significant threat to AI companies and their consumers. Our Generative AI Firewall provides a safety net against these failure modes.
At Robust Intelligence, we have built a multidisciplinary team of ML Engineers, AI security experts, and software engineers to advance the state of AI security. Together, we're building the future of secure, trustworthy AI.
About The Role
Are you passionate about the future of AI and committed to ensuring its safety and security? We're seeking Research Scientists to pioneer our efforts in tracking and mitigating emerging threats to AI systems.
In this role, you will:
Track and Analyze Threats: Monitor and analyze emerging threats to AI/ML models, applications, and environments.Deep Understanding: Possess expertise in foundation models like Large Language Models and diffusion models.Develop AI Firewall: Develop and implement strategies to detect and mitigate identified threats, including prototyping innovative approaches.Lead Red-Teaming Exercises: Lead red-teaming exercises and vulnerability assessments for generative AI technologies, addressing safety and security vulnerabilities.Publish Insights: Author blog posts, white papers, or research papers on emerging threats in AI safety and security.Collaborate and Innovate: Collaborate with cross-functional teams to translate research into product features and shape our machine learning culture.
Required Qualifications:
Undergraduate degree in EECS, Math, or Physics.Strong programming skills in Python and deep knowledge of machine learning tools like PyTorch.Strong algorithmic and problem-solving skills.Fluency in reading academic papers on AI/ML.
Preferred Qualifications:
Graduate degree in ML or related field with an established research record.Experience in AI/ML safety and security risks (e.g., data poisoning, adversarial attacks).Experience developing proof-of-concept exploits for new attacks.
